When michael debellis and his colleagues at the university of pittsburg school of medicine failed to show a statistically significant difference in volume between 44 maltreated children with ptsd and 61 healthy controls what was the significance of the outlier study?